When center-symmetric local binary pattern (CS-LBP) in pedestrian detection is extracted,the central pixel in the neighborhood is not involved in the calculation,and the art ificial judgment threshold is subjective,so different sub blocks fail to be dist inguishet.We propose a pedestrian detection algorithm based on adaptive HCS-LBP (Haar-like CS-LBP) features. The algorithm first constructs an HCS-LBP feature coding method,and uses local central symmetric mode to reduce the encoding length. The integral image method is used to quickly calculate the time complexity,t he global adaptive threshold is calculated by introducing the gray level probability into the central pixel,and the local adaptive thresho ld of neighborhood pixels is obtained by using the Gauss matrix to highlight the objective texture information of the image.Then the cen ter pixel is involved in the image coding,and the weights of different sub blocks are determined through the information entropy,so as to enhance the description ability of the image features. Finally,histogram intersection kernel SVM (HIKSVM) is used to train the samples to improve the accuracy of the classifier.The experimental results in the INRIA database show that the algorithm can effective ly improve the utilization of image information and improve the performance of the pedestrian detection algorithm.The accuracy rate is 0.4%~7.8% higher than that in other related literatures.